Variational level set segmentation for the human motion capture analysis

In Image processing and computer vision, analyzing and compressing the human motion capture data is a compelling challenge that is a trial with many researchers. For addressing this challenge, the variational level set method is used for segmenting the human motion from experimental motion capture data. First, the Mocap (Motion Capture) data is separated into aligned motion sequences. After that, Gabor filter is applied for the feature extraction prior to dimensionality reduction using the Principal Component Analysis (PCA) method. Then the boundary of the motion data is tracked by minimizing energy functional using the level set method. The results of the segmentation are found to be good and accurate. To verify the effectiveness of this method, Gabor filter is compared with Gaussian filter and the result depicted that Gabor filter greatly reduces the processing time.